Chris J. Needham

A primer on learning in Bayesian networks for computational biology

Bayesian networks (BNs) provide a neat and compact representation for expressing joint probability distributions (JPDs) and for inference. They are becoming increasingly important in the biological sciences for the tasks of inferring cellular networks [1], modelling protein signalling pathways [2], systems biology, data integration [3], classification [4], and genetic data analysis [5]. The representation and use of probability theory makes BNs suitable for combining domain knowledge and data, expressing causal relationships, avoiding overfitting a model to training data, and learning from incomplete datasets. The probabilistic formalism provides a natural treatment for the stochastic nature of biological systems and measurements. This primer aims to introduce BNs to the computational biologist, focusing on the concepts behind methods for learning the parameters and structure of models, at a time when they are becoming the machine learning method of choice. There are many applications in biology where we wish to classify data; for example, gene function prediction. To solve such problems, a set of rules are required that can be used for prediction, but often such knowledge is unavailable, or in practice there turn out to be many exceptions to the rules or so many rules that this approach produces poor results. Machine learning approaches often produce better results, where a large number of examples (the training set) is used to adapt the parameters of a model that can then be used for performing predictions or classifications on data. There are many different types of models that may be required and many different approaches to training the models, each with its pros and cons. An excellent overview of the topic can be found in [6] and [7]. Neural networks, for example, are often able to learn a model from training data, but it is often difficult to extract information about the model, which with other methods can provide valuable insights into the data or problem being solved. A common problem in machine learning is overfitting, where the learned model is too complex and generalises poorly to unseen data. Increasing the size of the training dataset may reduce this; however, this assumes more training data is readily available, which is often not the case. In addition, often it is important to determine the uncertainty in the learned model parameters or even in the choice of model. This primer focuses on the use of BNs, which offer a solution to these issues. The use of Bayesian probability theory provides mechanisms for describing uncertainty and for adapting the number of parameters to the size of the data. Using a graphical representation provides a simple way to visualise the structure of a model. Inspection of models can provide valuable insights into the properties of the data and allow new models to be produced.

Tracking multiple sports players through occlusion, congestion and scale

Tracking sports players over a large playing area is a challenging problem. The players move quickly, and have large variations in their silhouettes. This paper presents a framework for multi-object tracking, using a CONDENSATION based approach. Each player being tracked is independently fitted to a model, and the sampling probability for the group of samples is calculated as a function of the fitness score of each player. This function rewards consistently good scores, but punishes a group of some very good and some very bad fitness scores. Ground plane information is used throughout, and the predictive stage of the algorithm is improved to incorporate estimates of position from Kalman filters. This helps group the estimated positions of each player, and to aid in tracking through occlusions.

Inference in Bayesian networks

Bayesian networks are increasingly important for integrating biological data and for inferring cellular networks and pathways. What are Bayesian networks and how are they used for inference?

Performance evaluation metrics and statistics for positional tracker evaluation

This paper discusses methods behind tracker evaluation, the aim being to evaluate how well a tracker is able to determine the position of a target object. Few metrics exist for positional tracker evaluation; here the fundamental issues of trajectory comparison are addressed, and metrics are presented which allow the key features to be described. Often little evaluation on how precisely a target is tracked is presented in the literature, with results detailing for what percentage of the time the target was tracked. This issue is now emerging as a key aspect of tracker performance evaluation. The metrics developed are applied to real trajectories for positional tracker evaluation. Data obtained from a sports player tracker on video of a 5-a-side soccer game, and from a vehicle tracker, is analysed. These give quantitative positional evaluation of the performance of computer vision tracking systems, and provides a framework for comparison of different methods and systems on benchmark data sets.

Protocols from perceptual observations

This paper presents a cognitive vision system capable of autonomously learning protocols from perceptual observations of dynamic scenes. The work is motivated by the aim of creating a synthetic agent that can observe a scene containing interactions between unknown objects and agents, and learn models of these sufficient to act in accordance with the implicit protocols present in the scene. Discrete concepts (utterances and object properties), and temporal protocols involving these concepts, are learned in an unsupervised manner from continuous sensor input alone. Crucial to this learning process are methods for spatio-temporal attention applied to the audio and visual sensor data. These identify subsets of the sensor data relating to discrete concepts. Clustering within continuous feature spaces is used to learn object property and utterance models from processed sensor data, forming a symbolic description. The progol Inductive Logic Programming system is subsequently used to learn symbolic models of the temporal protocols presented in the presence of noise and over-representation in the symbolic data input to it. The models learned are used to drive a synthetic agent that can interact with the world in a semi-natural way. The system has been evaluated in the domain of table-top game playing and has been shown to be successful at learning protocol behaviours in such real-world audio-visual environments.

Predicting the effect of missense mutations on protein function: analysis with Bayesian networks

BackgroundA number of methods that use both protein structural and evolutionary information are available to predict the functional consequences of missense mutations. However, many of these methods break down if either one of the two types of data are missing. Furthermore, there is a lack of rigorous assessment of how important the different factors are to prediction.ResultsHere we use Bayesian networks to predict whether or not a missense mutation will affect the function of the protein. Bayesian networks provide a concise representation for inferring models from data, and are known to generalise well to new data. More importantly, they can handle the noisy, incomplete and uncertain nature of biological data. Our Bayesian network achieved comparable performance with previous machine learning methods. The predictive performance of learned model structures was no better than a naïve Bayes classifier. However, analysis of the posterior distribution of model structures allows biologically meaningful interpretation of relationships between the input variables.ConclusionThe ability of the Bayesian network to make predictions when only structural or evolutionary data was observed allowed us to conclude that structural information is a significantly better predictor of the functional consequences of a missense mutation than evolutionary information, for the dataset used. Analysis of the posterior distribution of model structures revealed that the top three strongest connections with the class node all involved structural nodes. With this in mind, we derived a simplified Bayesian network that used just these three structural descriptors, with comparable performance to that of an all node network.

Cognitive Vision: Integrating Symbolic Qualitative Representations with Computer Vision

We describe the challenge of combining continuous computer vision techniques and qualitative, symbolic methods to achieve a system capable of cognitive vision. Key to a truly cognitive system, is the ability to learn: to be able to build and use models constructed autonomously from sensory input. In this paper we overview a number of steps we have taken along the route to the construction of such a system, and discuss some remaining challenges.

Combining the interactome and deleterious SNP predictions to improve disease gene identification

A method has been developed for the prediction of proteins involved in genetic disorders. This involved combining deleterious SNP prediction with a system based on protein interactions and phenotype distances; this is the first time that deleterious SNP prediction has been used to make predictions across linkage‐intervals. At each step we tested and selected the best procedure, revealing that the computationally expensive method of assigning medical meta‐terms to create a phenotype distance matrix was outperformed by a simple word counting technique. We carried out in‐depth benchmarking with increasingly stringent data sets, reaching precision values of up to 75% (19% recall) for 10‐Mb linkage‐intervals (averaging 100 genes). For the most stringent (worst‐case) data we attained an overall recall of 6%, yet still achieved precision values of up to 90% (4% recall). At all levels of stringency and precision the addition of predicted deleterious SNPs was shown to increase recall. Hum Mutat 0, 1–9, 2009.

Inductive learning spatial attention

This paper investigates the automatic induction of spatial attention from the visual observation of objects manipulated on a table top. In this work, space is represented in terms of a novel observer-object relative reference system, named Local Cardinal System, defined upon the local neighbourhood of objects on the table. We present results of applying the proposed methodology on five distinct scenarios involving the construction of spatial patterns of coloured blocks.

Multi-resolution template kernels

Domains in which shapes of objects change rapidly and significantly are a challenge for existing representation techniques: sport is a good example of this. We present a texture-based approach that copes with these problems in addition to resolution variation. A set of exemplar poses are learned from subsampled example images of the target object, creating a set of multi-resolution template kernels which when convolved with the image respond suitably. This technique may then be used in established tracking algorithms (e.g. CONDENSATION [Isard, M et al., 1996]). We demonstrate the technique in two domains, and suggest a Markov approach using it to model behaviour.